Lucas M, O’Reilly EJ, Pan A, et al. Coffee, caffeine, and risk of completed suicide: Results from three prospective cohorts of American adults. World J Biol Psychiatry. 2014;15(5):377-386.
The investigators explored the relationship between coffee/caffeine consumption and completed suicide.
Observational study utilizing data collected from 3 prospective long-term longitudinal cohort studies: the Health Professionals Follow-up Study; the Nurses’ Health Study; and the Nurses’ Health Study II
After excluding individuals with diagnoses of cardiovascular disease or cancer at baseline, the study included a total of 43,599 men and 164,825 women: 43,599 men enrolled in the Health Professionals Follow-up Study (HPFS, 1988-2008); 73,820 women enrolled in the Nurses’ Health Study (NHS, 1992-2008); and another 91,005 women enrolled the Nurses’ Health Study II (NHS II, 1993-2007). Men enrolled in the HPFS were between 40 and 75 years of age in 1986, women enrolled in the NHS were between 30 and 55 years old in 1976, and the women in the NHS II were between 25 and 42 years old in 1989. All participants were registered nurses or other healthcare professionals. Overall, NHS II participants were younger and more than twice as likely to use antidepressants compared to participants in other 2 cohort studies.
Study Parameters Assessed
Participants in all 3 cohorts completed detailed food-frequency questionnaires every 2 to 4 years. Relevant beverages and foods on the questionnaires included coffee, decaffeinated coffee, tea (“nonherbal”), carbonated soda (with or without caffeine), and chocolate. Participants indicated consumption of a serving (eg, 8 fl oz=1 c coffee) at a frequency ranging from never to 6 or more servings daily. Caffeine content was assumed to be 137 mg per cup of coffee, 47 mg per cup of tea, 46 mg per can of soda, and 7 mg per serving of chocolate.
Primary Outcome Measures
Cause of death for was determined from death certificates using International Classification of Diseases, 8th Edition (ICD-8) codes E950 to E959 to identify all cases of suicide and self-inflicted injuries.
Coffee was the largest dietary contributor to total caffeine intake, comprising 71% to 80% of caffeine consumed. Mean daily caffeine intake was 218 mg, 169 mg, and 186 mg for the NHS, NHS II, and HPFS cohorts, respectively. A total of 277 completed suicides were documented: 47 in the NHS, 66 in the NHS II, and 164 in the HPFS.
As tobacco and exogenous estrogens (eg, hormonal contraceptives) are known to significantly alter caffeine’s half-life in the body, effect modification by these factors was also examined. Alcohol use was examined as well. None of these factors was found to have significant effect on suicide risk.
Investigators adjusted for smoking, alcohol, antidepressant use, and marital/partnership status. Of these, tobacco use proved to be a negative confounder. After adjustment, a significant trend of decreasing relative risk (RR) for completed suicide was noted with increased coffee consumption (P-value for trend=0.02-0.04 depending upon the study). Pooled data from all 3 cohorts showed a RR for suicide was 0.55 (95% confidence interval [CI]: 0.38-0.78) for those consuming 2 to 3 cups per day and 0.47 (95% CI:0.27-0.81) for those consuming 4 or more cups per day compared to those consuming 1 cup or fewer per day. The P-value for the trend was highly significant (P for trend<0.001).
Results for caffeine were mixed. For the pooled multivariate analysis, the RR of suicide was 0.54 (95% CI:0.30- 0.94) for daily intake of 400 mg to 550 mg, and 0.63 (95% CI:0.39-1.04) for 550 mg or more per day compared to less than 100 mg per day (P for trend=0.005). In the individual cohorts, the computed P-value for the trend was significant (0.05) only for the NHS data.
No association with suicide risk was noted for either decaffeinated coffee or tea consumption.
The findings of this study suggest that suicide risk decreases in a dose-dependent manner with increasing coffee consumption. As caffeine has a similar, if lesser protective effect, Lucas et al suggest it may be a large contributor to the observed effect of coffee. The investigators note that the low reported consumption of decaffeinated coffee limits the ability to fully explore this possibility. Similarly, low reported consumption of tea limits ability to explore the effects of tea on suicide risk.
Although Lucas et al focused largely upon caffeine as a key contributor to the observed effect, coffee contains a rich array of phytochemicals beyond caffeine, with their presence or absence determined in part by roasting and preparation methods. These polyphenols and other compounds have been also been independently associated with positive health effects.1 Mechanistically, these phytochemicals may modulate caffeine’s effects or directly impact physiological health in some other way.2
Lucas et al identified 2 other cohort studies showing similar associations between coffee/caffeine consumption and suicide risk: the Northern California Kaiser Permanente study3 and the NHS 10-year follow-up study.4 A third cohort study conducted in Finland showed a J-shaped association between coffee consumption and suicide risk. In this study, suicide risk decreased with coffee consumption between 2 to 3 cups per day and 6 to 7 cups per day, compared to 1 cup or fewer per day. However, suicide risk increased for those reporting 8 to 9 cups per day and 10 cups or more per day.5 As the highest coffee intake category for the present study was 6 or more cups per day, it is not possible to identify whether the current study would have shown a similar J-shaped curve at higher intakes. Noting that mentally ill individuals sometimes self-medicate with caffeine,6,7 Lucas et al suggest in the present study that very high doses of coffee might reflect an (ultimately insufficient) attempt to alleviate dysphoria or other symptoms. It is also possible that at very high intake, adverse effects of caffeine may begin to outweigh any positive ones.
Caffeine’s described effects on the central nervous system are complex. Overall, its effects are arousal and stimulation. Dopaminergic transmission is enhanced8 with increased turnover of serotonin, dopamine, and norepinephrine, much of this mediated by adenosine (A2a and A1) receptor antagonism.8-10 As monoamine deficiency is considered an underlying factor in depression,11 it is plausible that caffeine may possess antidepressant activity. Indeed, epidemiological studies have associated caffeine intake with lower risk of depression.12,13 However, caffeine may also aggravate anxiety, insomnia, or panic attacks.14,15
Tolerance develops to caffeine’s effects, and a withdrawal syndrome has been described. It is recognized as a drug of abuse by both the Diagnostic and Statistical Manual of Mental Disorder, 5th Edition and ICD-10.15
Caffeine is metabolized by cytochrome P450 1A2 (CYP1A2) and thus is subject to interaction with drugs and other substances that alter CYP1A2 activity level. Polycyclic aromatic hydrocarbons and other compounds found in tobacco smoke may decrease the half-life by as much as 50% via enzyme induction. Conversely, exogenous estrogens may double caffeine’s half-life by inhibiting 1A2 activity.16,17 Of greater concern, caffeine may interact with multiple psychiatric medications. Selective serotonin reuptake inhibitors (SSRIs), especially fluvoxamine, also inhibit CYP1A2. In addition, several other psychiatric medications, including antipsychotics, benzodiazepines, and tricyclic antidepressants are metabolized at least in part by this enzyme.17 Through a separate mechanism, caffeine may increase lithium clearance through the kidneys, leading to reduced plasma levels and raising the possibility of treatment failure.15
The clinical implications of these interactions are significant. For example, a patient initiating SSRI treatment may begin experiencing adverse effects from caffeine (eg, anxiety, insomnia, or even panic attacks). If the patient and/or provider are unaware of the potential interaction between caffeine and SSRIs, the symptoms may be ascribed to the SSRI rather than to the interaction. Minimizing or avoiding caffeine may improve tolerance of the SSRI. Alternatively, a patient taking lithium may experience treatment failure if caffeine intake is increased without dose adjustment (conversely, lithium toxicity may occur if caffeine is discontinued without dose adjustment). In general, it is best to monitor lithium levels closely when a patient is making significant dietary changes.
As with all research, consideration of study outcomes must include notation of the study’s limitations. First, the present study lacked power to adequately evaluate the effect of decaffeinated coffee and tea consumption on suicide risk. The same applies to exploration of antidepressant use, as the collected data lacked specifics such as dosage and duration of use. Also, the coffee intake data collected failed to differentiate between high and very high consumption of coffee/caffeine. As a result, this study does not provide further insight into the J-shaped association between suicide risk and coffee consumption observed by Tanskanen et al.5 In addition, participants were predominantly non-Hispanic white health professionals who may differ in key ways from the general population. Finally, the analysis excluded patients with previously diagnosed cardiovascular disease or cancer.
This study adds to the growing body of evidence demonstrating how complex the health effects of coffee and similar beverages and/or foods are. The suggestion that moderate consumption of coffee may reduce suicide risk is intriguing. It is also possible, however, that people are more likely to drink coffee when less depressed. However, recommending coffee or an increase in caffeine intake on the basis of possible antidepressant activity is premature. Lucas et al point out that many people self-regulate their coffee and caffeine intake for optimal effects.
Long-standing naturopathic traditions hold that coffee and by extension caffeine should be minimized or eliminated if one is to optimize health. The founders of naturopathic medicine observed that patients improved with caffeine avoidance. As scientific research reveals more benefits of coffee, tea, and chocolate, perhaps it is time to revisit tradition. Perhaps the positive effects observed by the founders were related to other changes, such as increased water intake (to replace the coffee, tea, etc). Patients presenting for naturopathic care may expect to be advised to cut back or eliminate coffee and/or caffeine. A better approach may be to consider intake in the context of the patient’s overall health picture. If the patient reports symptoms (eg, anxiety, insomnia) that could be alleviated by moderation of coffee or caffeine, reduced intake is appropriate. However, in the absence of such symptoms, other treatments may provide greater benefit. In addition, when working with mental health patients, evaluating the patient’s stability and whether minimizing or eliminating caffeine may trigger symptom exacerbation is critical.
The list of health benefits ascribed to coffee and similar beverages and/or foods is long and growing. It may reduce risk for neurological conditions such as Parkinson’s and Alzheimer’s diseases while also having beneficial effects on health conditions of the body: multiple types of cancer, diabetes, cardiovascular diseases, and liver function.2,18-20 For individuals able to tolerate them in moderate amounts, continued enjoyment of these beverages and foods may be encouraged.
Patients presenting for naturopathic care may expect to be advised to cut back or eliminate coffee and/or caffeine. A better approach may be to consider intake in the context of the patient’s overall health picture.
In this study utilizing data collected from 3 prospective, long-term studies, researchers found that coffee consumption was inversely correlated to completed suicide. Naturopathic doctors have advised against the use of caffeine for decades; will this study change our minds about the benefits of coffee in the lives of our patients with mood disorders?